AI is more accurate than doctors in emergency diagnoses - study

"The AI system identified the correct or nearly correct diagnosis in 67% of cases, compared to 50% to 55% percent for doctors."

"When more detailed patient data was available, the AI's accuracy increased to 82%, while the doctors' accuracy ranged from 70% to 79%."

"In a test using clinical cases, the AI achieved 89% accuracy, while doctors using traditional tools such as search engines reached 34%."

A study from Harvard Medical School reveals that AI can surpass doctors in emergency care triage assessments. In initial evaluations, AI achieved a correct diagnosis in 67% of cases, while doctors' accuracy ranged from 50% to 55%. With more detailed patient data, AI's accuracy rose to 82%, compared to 70% to 79% for doctors. Additionally, in developing treatment plans, AI reached 89% accuracy, significantly outperforming doctors who achieved only 34% using traditional search tools.

Read at Computerworld